Using an experimental factorial design we have recently demonstrated significant differences between primary care providers in the US and the UK in the clinical management of common medical problems of the elderly (coronary heart disease and depression). These differences are highly significant (most p<.001 ) and unconfounded by patient attributes or physician characteristics. This Competing Continuation will determine whether these exciting health care system differences in clinical decision-making (CDM) are sustained in a third, quite different system (Germany's federalized and corporatized system). Research on health system contributions to variations in CDM obviously requires inclusion of several systems for comparative purposes. Replication of our experiment in a different national setting (Germany) will add scientific confidence to recent findings of system differences in the US and the UK. This research represents a paradigm shift in studies of CDM, by moving the focus from patient attributes (prescriptive CDM) and provider characteristics (descriptive CDM) to a more sociological "third generation" approach (health care system contributions) and to "fourth generation" studies of underlying cognitive (and reasoning) processes. Novel use of a classical experimental (factorial) design (as opposed to observational data) permits unconfounded estimation of different effects. It builds cost-efficiently on work already completed on time and within budget. It has policy implications at the level of patient care, provider training and the organization and financing of medical care. It offers a new explanation for observed international variations in disease rates. The most recent issue of AJPH (February 2003) highlights the urgent need for comparative analyses of different health systems in order to provide policy lessons for the rapidly changing US health care system.